Advertisement

Deep Learning of Renal Scans in Children with Antenatal Hydronephrosis

Published:January 05, 2023DOI:https://doi.org/10.1016/j.jpurol.2022.12.017

      Summary

      Introduction

      Antenatal hydronephrosis (ANH) is one of the most common anomalies identified on prenatal ultrasound, found in up to 4.5% of all pregnancies. Children with ANH are surveilled with repeated renal ultrasound and when there is high suspicion for a ureteropelvic junction obstruction on renal ultrasound, a mercaptuacetyltriglycerine (MAG3) Lasix renal scan is performed to evaluate for obstruction. However, the challenging interpretation of MAG3 renal scans places patients at risk of misdiagnosis.

      Objective

      Our objective was to analyze MAG3 renal scans using machine learning to predict renal complications. We hypothesized that our deep learning model would extract features from MAG3 renal scans that can predict renal complications in children with ANH.

      Study Design

      We performed a case-control study of MAG3 studies drawn from a population of children with ANH concerning for ureteropelvic junction obstruction evaluated at our institution from January 2009 until June of 2021. The outcome was renal complications that occur ≥ 6 months after an equivocal MAG-3 renal scan. We created two machine learning models: a deep learning model using the radiotracer concentration versus time data from the kidney of interest and a random forest model created using clinical data. The performance of the models was assessed using measures of diagnostic accuracy.

      Results

      We identified 152 eligible patients with available images of which 62 were cases and 90 were controls. The deep learning model predicted future renal complications with an overall accuracy of 73% (95% confidence inteveral [CI] 68-76%) and an AUC of 0.78 (95% CI 0.7, 0.84). The random forest model had an accuracy of 62% (95% CI 60-66%) and an AUC of 0.67 (95% CI. 0 64, 0.72)

      Discussion

      Our deep learning model predicted patients at high risk of developing renal complications following an equivocal renal scan and discriminate those at low risk with moderately high accuracy (73%). The deep learning model outperformed the clinical model built from clinical features classically used by urologists for surgical decision making.

      Conclusion

      Our models have the potential to influence clinical decision making by providing supplemental analytical data from MAG3 scans that would not otherwise be available to urologists. Future multi-institutional retrospective and prospective trials are needed to validate our model.
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Journal of Pediatric Urology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Ismaili K.
        • Hall M.
        • Donner C.
        • et al.
        Results of systematic screening for minor degrees of fetal renal pelvis dilatation in an unselected population.
        Am J Obstet Gynecol. 2003; 188: 242-246https://doi.org/10.1067/mob.2003.81
        • Dudley J.A.
        • Haworth J.M.
        • McGraw M.E.
        • Frank J.D.
        • Tizard E.J.
        Clinical relevance and implications of antenatal hydronephrosis.
        Arch Dis Child Fetal Neonatal Ed. 1997; 76: F31-F34https://doi.org/10.1136/fn.76.1.f31
        • Chiodini B.
        • Ghassemi M.
        • Khelif K.
        • Ismaili K.
        Clinical Outcome of Children With Antenatally Diagnosed Hydronephrosis.
        Front Pediatr. 2019; 7: 103https://doi.org/10.3389/fped.2019.00103
        • Becker A.
        • Baum M.
        Obstructive uropathy.
        Early Hum Dev. 2006; 82: 15-22https://doi.org/10.1016/j.earlhumdev.2005.11.002
        • Durand E.
        • Blaufox M.D.
        • Britton K.E.
        • et al.
        International Scientific Committee of Radionuclides in Nephrourology (ISCORN) consensus on renal transit time measurements.
        Semin Nucl Med. 2008; 38: 82-102https://doi.org/10.1053/j.semnuclmed.2007.09.009
        • Kletter K.
        • Nürnberger N.
        Diagnostic potential of diuresis renography: limitations by the severity of hydronephrosis and by impairment of renal function.
        Nucl Med Commun. 1989; 10: 51-61
        • Karacalioglu O.
        • Ilgan S.
        • Arslan N.
        • Emer O.
        • Ozguven M.
        Unilateral temporary functional stasis in the upper urinary tract caused by “a filled bladder” on Tc-99m DTPA diuresis renography: a teaching case.
        Ann Nucl Med. 2005; 19: 511-514https://doi.org/10.1007/BF02985580
        • Ulman I.
        • Jayanthi V.R.
        • Koff S.A.
        The long-term followup of newborns with severe unilateral hydronephrosis initially treated nonoperatively.
        J Urol. 2000; 164: 1101-1105https://doi.org/10.1097/00005392-200009020-00046
        • Yin S.
        • Peng Q.
        • Li H.
        • et al.
        Computer-Aided Diagnosis of Congenital Abnormalities of the Kidney and Urinary Tract in Children Using a Multi-Instance Deep Learning Method Based on Ultrasound Imaging Data.
        Proc IEEE Int Symp Biomed Imaging. 2020; 2020: 1347-1350https://doi.org/10.1109/isbi45749.2020.9098506
        • Weaver J.K.
        • Milford K.
        • Rickard M.
        • et al.
        Deep learning imaging features derived from kidney ultrasounds predict chronic kidney disease progression in children with posterior urethral valves.
        Pediatr Nephrol. 2022; (Published online July 22)https://doi.org/10.1007/s00467-022-05677-0
      1. Yin S. Multi-instance Deep Learning with Graph Convolutional Neural Networks for Diagnosis of Kidney Diseases Using Ultrasound Imaging. Greenspan H et al (eds) Uncertainty for Safe Utilization of Machine Learning in Medical Imaging and Clinical Image-Based Procedures CLIP 2019, UNSURE 2019 Lecture Notes in Computer Science. 2019;11840.

        • Smokvina A.
        • Grbac-Ivanković S.
        • Girotto N.
        • Dezulović M.S.
        • Saina G.
        • Barković M.M.
        The renal parenchyma evaluation: MAG3 vs. DMSA.
        Coll Antropol. 2005; 29: 649-654
        • Ritchie G.
        • Wilkinson A.G.
        • Prescott R.J.
        Comparison of differential renal function using technetium-99m mercaptoacetyltriglycine (MAG3) and technetium-99m dimercaptosuccinic acid (DMSA) renography in a paediatric population.
        Pediatr Radiol. 2008; 38: 857-862https://doi.org/10.1007/s00247-008-0908-8
        • Nimmo M.
        Measurement of relative renal function. A comparison of methods and assessment of reproducibility.
        British Journal of Radiology. 1987; 60: 861-864
        • Calle-Toro J.S.
        • Barrera C.A.
        • Khrichenko D.
        • Otero H.J.
        • Serai S.D.
        R2 relaxometry based MR imaging for estimation of liver iron content: A comparison between two methods.
        Abdom Radiol (NY). 2019; 44: 3058-3068https://doi.org/10.1007/s00261-019-02074-4
        • Krill A.J.
        • Varda B.K.
        • Freidberg N.A.
        • et al.
        Predicting the likelihood of prolongation of half-time among infants with initially indeterminate drainage values: A single-institution retrospective study of 535 patients with ureteropelvic junction obstruction.
        J Pediatr Urol. 2021; 17: 512.e1-512.e7https://doi.org/10.1016/j.jpurol.2021.05.022
        • Conway J.J.
        • Maizels M.
        The “well tempered” diuretic renogram: a standard method to examine the asymptomatic neonate with hydronephrosis or hydroureteronephrosis. A report from combined meetings of The Society for Fetal Urology and members of The Pediatric Nuclear Medicine Council--The Society of Nuclear Medicine.
        J Nucl Med. 1992; 33: 2047-2051
        • Peters C.A.
        • Carr M.C.
        • Lais A.
        • Retik A.B.
        • Mandell J.
        The response of the fetal kidney to obstruction.
        J Urol. 1992; 148: 503-509https://doi.org/10.1016/s0022-5347(17)36640-5
        • Donoso G.
        • Ham H.
        • Tondeur M.
        • Piepsz A.
        Influence of early furosemide injection on the split renal function.
        Nucl Med Commun. 2003; 24: 791-795https://doi.org/10.1097/01.mnm.0000080253.50447.94
        • Dhillon H.K.
        Prenatally diagnosed hydronephrosis: the Great Ormond Street experience.
        Br J Urol. 1998; 81: 39-44https://doi.org/10.1046/j.1464-410x.1998.0810s2039.x
        • Shamshirsaz A.A.
        • Ravangard S.F.
        • Egan J.F.
        • et al.
        Fetal hydronephrosis as a predictor of neonatal urologic outcomes.
        J Ultrasound Med. 2012; 31: 947-954https://doi.org/10.7863/jum.2012.31.6.947
        • Dias C.S.
        • Silva J.M.P.
        • Pereira A.K.
        • et al.
        Diagnostic accuracy of renal pelvic dilatation for detecting surgically managed ureteropelvic junction obstruction.
        J Urol. 2013; 190: 661-666https://doi.org/10.1016/j.juro.2013.02.014
        • Coplen D.E.
        Prenatal intervention for hydronephrosis.
        J Urol. 1997; 157: 2270-2277
        • Melo F.F.
        • Mak R.H.
        • Simões E Silva A.C.
        • et al.
        Evaluation of Urinary Tract Dilation Classification System for Prediction of Long-Term Outcomes in Isolated Antenatal Hydronephrosis: A Cohort Study.
        J Urol. 2021; 206: 1022-1030https://doi.org/10.1097/JU.0000000000001899
        • Nie D.
        • Zhang H.
        • Adeli E.
        • Liu L.
        • Shen D.
        3D Deep Learning for Multi-modal Imaging-Guided Survival Time Prediction of Brain Tumor Patients.
        Med Image Comput Comput Assist Interv. 2016; 9901: 212-220https://doi.org/10.1007/978-3-319-46723-8_25
        • Esteva A.
        • Kuprel B.
        • Novoa R.A.
        • et al.
        Dermatologist-level classification of skin cancer with deep neural networks.
        Nature. 2017; 542: 115-118https://doi.org/10.1038/nature21056
        • Rajkomar A.
        • Lingam S.
        • Taylor A.G.
        • Blum M.
        • Mongan J.
        High-Throughput Classification of Radiographs Using Deep Convolutional Neural Networks.
        J Digit Imaging. 2017; 30: 95-101https://doi.org/10.1007/s10278-016-9914-9
        • Litjens G.
        • Sánchez C.I.
        • Timofeeva N.
        • et al.
        Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis.
        Sci Rep. 2016; 6: 26286https://doi.org/10.1038/srep26286
        • Gulshan V.
        • Peng L.
        • Coram M.
        • et al.
        Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs.
        JAMA. 2016; 316: 2402-2410https://doi.org/10.1001/jama.2016.17216
        • Cheng J.Z.
        • Ni D.
        • Chou Y.H.
        • et al.
        Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans.
        Sci Rep. 2016; 6: 24454https://doi.org/10.1038/srep24454
        • Lee J.N.
        • Kang J.K.
        • Jeong S.Y.
        • et al.
        Predictive value of cortical transit time on MAG3 for surgery in antenatally detected unilateral hydronephrosis caused by ureteropelvic junction stenosis.
        J Pediatr Urol. 2018; 14: 55.e1-55.e6https://doi.org/10.1016/j.jpurol.2017.08.009
        • Piepsz A.
        • Tondeur M.
        • Ham H.
        NORA: a simple and reliable parameter for estimating renal output with or without frusemide challenge.
        Nucl Med Commun. 2000; 21: 317-323https://doi.org/10.1097/00006231-200004000-00005
        • Acker M.R.
        • Clark R.
        • Anderson P.
        Gravity-assisted drainage imaging in the assessment of pediatric hydronephrosis.
        Can Urol Assoc J. 2016; 10: 96-100https://doi.org/10.5489/cuaj.3237